3 files changed, 40 insertions, 7 deletions

diff --git a/lib/Analysis/CodeMetrics.cpp b/lib/Analysis/CodeMetrics.cpp
index 84b3266..401c1be 100644
--- a/lib/Analysis/CodeMetrics.cpp
+++ b/lib/Analysis/CodeMetrics.cpp
@@ -165,6 +165,14 @@ void CodeMetrics::analyzeBasicBlock(const BasicBlock *BB,
     if (isa<ExtractElementInst>(II) || II->getType()->isVectorTy())
       ++NumVectorInsts;
 
+    if (const CallInst *CI = dyn_cast<CallInst>(II))
+      if (CI->hasFnAttr(Attribute::NoDuplicate))
+        notDuplicatable = true;
+
+    if (const InvokeInst *InvI = dyn_cast<InvokeInst>(II))
+      if (InvI->hasFnAttr(Attribute::NoDuplicate))
+        notDuplicatable = true;
+
     ++NumInsts;
   }
 
@@ -182,8 +190,7 @@ void CodeMetrics::analyzeBasicBlock(const BasicBlock *BB,
   // if someone is using a blockaddress without an indirectbr, and that
   // reference somehow ends up in another function or global, we probably
   // don't want to inline this function.
-  if (isa<IndirectBrInst>(BB->getTerminator()))
-    containsIndirectBr = true;
+  notDuplicatable |= isa<IndirectBrInst>(BB->getTerminator());
 
   // Remember NumInsts for this BB.
   NumBBInsts[BB] = NumInsts - NumInstsBeforeThisBB;
diff --git a/lib/Analysis/InlineCost.cpp b/lib/Analysis/InlineCost.cpp
index f88cabb..de89fea 100644
--- a/lib/Analysis/InlineCost.cpp
+++ b/lib/Analysis/InlineCost.cpp
@@ -54,6 +54,8 @@ class CallAnalyzer : public InstVisitor<CallAnalyzer, bool> {
   bool IsRecursiveCall;
   bool ExposesReturnsTwice;
   bool HasDynamicAlloca;
+  bool ContainsNoDuplicateCall;
+
   /// Number of bytes allocated statically by the callee.
   uint64_t AllocatedSize;
   unsigned NumInstructions, NumVectorInstructions;
@@ -128,8 +130,8 @@ public:
   CallAnalyzer(const DataLayout *TD, Function &Callee, int Threshold)
     : TD(TD), F(Callee), Threshold(Threshold), Cost(0),
       IsCallerRecursive(false), IsRecursiveCall(false),
-      ExposesReturnsTwice(false), HasDynamicAlloca(false), AllocatedSize(0),
-      NumInstructions(0), NumVectorInstructions(0),
+      ExposesReturnsTwice(false), HasDynamicAlloca(false), ContainsNoDuplicateCall(false),
+      AllocatedSize(0), NumInstructions(0), NumVectorInstructions(0),
       FiftyPercentVectorBonus(0), TenPercentVectorBonus(0), VectorBonus(0),
       NumConstantArgs(0), NumConstantOffsetPtrArgs(0), NumAllocaArgs(0),
       NumConstantPtrCmps(0), NumConstantPtrDiffs(0),
@@ -615,6 +617,9 @@ bool CallAnalyzer::visitCallSite(CallSite CS) {
     ExposesReturnsTwice = true;
     return false;
   }
+  if (CS.isCall() &&
+      cast<CallInst>(CS.getInstruction())->hasFnAttr(Attribute::NoDuplicate))
+    ContainsNoDuplicateCall = true;
 
   if (IntrinsicInst *II = dyn_cast<IntrinsicInst>(CS.getInstruction())) {
     switch (II->getIntrinsicID()) {
@@ -842,7 +847,9 @@ bool CallAnalyzer::analyzeCall(CallSite CS) {
 
   // If there is only one call of the function, and it has internal linkage,
   // the cost of inlining it drops dramatically.
-  if (F.hasLocalLinkage() && F.hasOneUse() && &F == CS.getCalledFunction())
+  bool OnlyOneCallAndLocalLinkage = F.hasLocalLinkage() && F.hasOneUse() &&
+    &F == CS.getCalledFunction();
+  if (OnlyOneCallAndLocalLinkage)
     Cost += InlineConstants::LastCallToStaticBonus;
 
   // If the instruction after the call, or if the normal destination of the
@@ -1008,6 +1015,12 @@ bool CallAnalyzer::analyzeCall(CallSite CS) {
     }
   }
 
+  // If this is a noduplicate call, we can still inline as long as 
+  // inlining this would cause the removal of the caller (so the instruction
+  // is not actually duplicated, just moved).
+  if (!OnlyOneCallAndLocalLinkage && ContainsNoDuplicateCall)
+    return false;
+
   Threshold += VectorBonus;
 
   return Cost < Threshold;
@@ -1025,6 +1038,7 @@ void CallAnalyzer::dump() {
   DEBUG_PRINT_STAT(NumInstructionsSimplified);
   DEBUG_PRINT_STAT(SROACostSavings);
   DEBUG_PRINT_STAT(SROACostSavingsLost);
+  DEBUG_PRINT_STAT(ContainsNoDuplicateCall);
 #undef DEBUG_PRINT_STAT
 }
 #endif
diff --git a/lib/Analysis/LoopInfo.cpp b/lib/Analysis/LoopInfo.cpp
index 1457bea..d88f8a8 100644
--- a/lib/Analysis/LoopInfo.cpp
+++ b/lib/Analysis/LoopInfo.cpp
@@ -213,10 +213,22 @@ bool Loop::isLoopSimplifyForm() const {
 /// isSafeToClone - Return true if the loop body is safe to clone in practice.
 /// Routines that reform the loop CFG and split edges often fail on indirectbr.
 bool Loop::isSafeToClone() const {
-  // Return false if any loop blocks contain indirectbrs.
+  // Return false if any loop blocks contain indirectbrs, or there are any calls
+  // to noduplicate functions.
   for (Loop::block_iterator I = block_begin(), E = block_end(); I != E; ++I) {
-    if (isa<IndirectBrInst>((*I)->getTerminator()))
+    if (isa<IndirectBrInst>((*I)->getTerminator())) {
       return false;
+    } else if (const InvokeInst *II = dyn_cast<InvokeInst>((*I)->getTerminator())) {
+      if (II->hasFnAttr(Attribute::NoDuplicate))
+        return false;
+    }
+
+    for (BasicBlock::iterator BI = (*I)->begin(), BE = (*I)->end(); BI != BE; ++BI) {
+      if (const CallInst *CI = dyn_cast<CallInst>(BI)) {
+        if (CI->hasFnAttr(Attribute::NoDuplicate))
+          return false;
+      }
+    }
   }
   return true;
 }